BAKER PERKINS IN THE CHOCOLATE AND CONFECTIONERY BUSINESS

As well as helping to put relevant business events into chronological context, attempting a history of Baker Perkins in a particular industry sector gives the opportunity to introduce information about how particular processes and technologies were developed and applied to the design of production equipment. The first part of this history covers the growth of Baker Perkins’ involvement in the chocolate and sugar confectionery industry, whilst the second describes, in some detail, the characteristics of the end-products and the challenges inherent in designing equipment to reproduce these under industrial production conditions.

In the early days, the "Perkins" side of the business tended to concentrate on bread baking with the "Baker" side concentrating mainly (but not exclusively) on biscuit and confectionery equipment. The history of Baker Perkins in the chocolate and confectionery industry, therefore, starts in 1890 at Willesden. For the history of the company before this date see:

The Origins of the Founders
History of Perkins, Bacon & Petch.
History of A.M. Perkins & Son Ltd.
History of Joseph Baker & Sons Ltd
History of Werner & Pfleiderer (London) Ltd.
History of Werner, Pfleiderer & Perkins Ltd.

It could perhaps be useful to start with a summary of the background to chocolate and sugar confectionery:

The Origins of Chocolate

The cacao tree has dozens of yellow-green pods hanging from its trunk and stems. The hand-sized fruits contain a fibrous white pulp, embedded in which are the dark, purple-coloured seeds that, after being dried and processed are recognisable as “chocolate beans”. Many people believe that the Aztecs first developed chocolate but chocolate goes back much farther. The ancient Maya, who inhabited what is now part of southern Mexico and Central America, certainly consumed chocolate. In fact, the word "cacao" is Mayan and some think chocolate may be even older. These early civilisations consumed chocolate as a bitter-tasting drink made of ground cacao beans mixed with a variety of local ingredients, wine or water, sometimes seasoned with vanilla, pimiento, and chilli pepper. It was thought to cure diarrhoea and dysentery, and was believed to be an aphrodisiac. Cacoa beans were also the local currency. In fact, in some parts of Central America, cacao beans were used as currency as recently as the last century.

The Spaniards introduced cacoa to Europe following their conquest of the Aztecs of Mexico and at first it was still being served as a beverage, but soon went through an important evolution: the chilli pepper was replaced by sugar. The new, sweetened, chocolate beverage was a luxury few could afford, but by the 17th century the drink was common among European nobility. However, in England, chocolate was more widely available, those who could afford it could enjoy chocolate drinks in the new coffee and chocolate houses of London.

The centre of the cacoa bean, known as the "nib," contains an average of 54 percent cocoa butter, a natural fat and in 1828, the Dutch chocolate maker Conrad J. van Houten patented an inexpensive method for pressing the fat from roasted cacao beans. His hydraulic press reduced the cocoa butter content by nearly half, creating a "cake" that could be pulverized into a fine powder known as "cocoa." The powder was then treated with alkaline salts (potassium or sodium carbonates) so that it would mix more easily with water. This not only made creating chocolate drinks much easier, but also made it possible to combine chocolate with sugar and then remix it with cocoa butter to create a solid. In 1849, English chocolate maker Joseph Storrs Fry produced what was arguably the world's first eating chocolate.

In the late 19th century, the Swiss developed a number of processes and two major developments occurred in 1879. Daniel Peter, a Swiss chocolate manufacturer, had the idea of using powdered milk (invented by Swiss Chemist Henri Nestle in 1867) to make a new kind of chocolate - milk chocolate, and Rudolphe Lindt invented a process called "conching", which greatly improved the quality of chocolate by making it more blendable.

Sugar Confectionery

Today, the sugar confectionery market sector covers a huge range of products - gums, jellies, chews, mints, boiled sweets, fudge, liquorice, toffees and caramels as well as chewing and bubblegum among others - but, like chocolate, the making of sugar confectionery (sweets, candy) can also be traced back for thousands of years to its origins in the East. The basic process was the simple boiling of sugar in a pan on an open fire and the first problem faced by those attempting to mechanise the process was to find a way of boiling the sugar other than on an open fire. Edward Metcalfe Shaw was the first man to do this and he had a long association with the Bakers. Using steam as his cooking medium, Shaw produced the "Eureka" machine, in which the sugar solution was pumped through a copper coil in a steam vessel – a basic principle used in all later continuous sugar cookers. His first machine was working before 1900 and within a few years he had produced a machine in which the cooked sugar emerged into a vacuum chamber, enabling not only cooking at a lower temperature but improvement of colour and dryness of the confection. By 1904, the company was claiming that 67 "Eureka" plants were at work in Great Britain, Canada, New Zealand, the United States, Australia, Europe, South Africa and Japan, the machine being the subject of 23 patents at home and abroad.

Baker Perkins involvement in the Chocolate and Sugar Confectionery business.

The Joseph Baker years

Joseph Baker & Sons Ltd. broke into the chocolate and confectionery industries soon after they moved into their new factory at Willesden in 1890 with a sifting and separating machine for sugar that was virtually a development of the original Baker flour sifter (See How it Started). The Bakers had seen their key task as gaining customers and had, for a number of years sold an increasing range of products connected with the food industry, issuing detailed catalogues at frequent intervals. The thirty-fourth edition had appeared by 1886, was said to look like a small pulpit bible and weighed five pounds. Anything that the food producer or grocer might need was on offer – including shop and bakery fittings, restaurant tables, a hand cart or bread van and even a Lancashire boiler! These products had a value beyond the modest profit derived from their sale, helping to bring in potential customers to their young, growing business. It comes as no surprise, therefore, that they met the opportunity for chocolate machinery by selling a range of equipment not all of their own manufacture.

Even in these early days, Joseph Baker & Sons was building strong relationships with some of the emerging companies who were to become giants of the industry in later years. The December 1964 issue of "Group News" records a close association with Cadburys going back to the nineteenth century:

“Since 1887, when Joseph Baker & Sons exhibited a wide range of chocolate processing equipment at the Royal Jubilee Exhibition in Manchester, there has been contact with Cadburys. Our earliest records show that in the closing years of the nineteenth century we supplied the company with a comprehensive range of chocolate and confectionery processing equipment. This link has been maintained as technical innovations have surpassed one another.

The first machine on record to be bought was a chocolate mixer in 1894 at the cost of £90 and this was followed later by chocolate coaters, tapping tables – for settling the chocolate moulds – and other equipment. By 1898, Cadburys had purchased six cream fondant depositors and five starch printers, which point to a thriving chocolate business even in those days.

By the beginning of the new century, wafer machines, fruit pulpers and cake machines had been supplied and, in 1904, Joseph Baker & Sons sold its very first five roll chocolate refiner to Cadburys. The rollers were of granite in those days but the machine is still the same in principle even today.

The year 1907 records the first caramel wrapping machine to go to Bournville, and form 1915 onwards there was a considerable delivery of chocolate melangeurs. In 1923, Baker Perkins Ltd sold Cadburys a chocolate piping and decorating machine. Later in the twenties, travelling ovens were regularly delivered as were Forgrove wrapping machines.

More recently, Baker Perkins Ltd have supplied a number of biscuit ovens and additional plant is on order to meet the expansion of the confectionery side of Cadbury’s business”.

At the 1900 Paris Exposition, Joseph Baker & Sons was showing a complete working "Chocolaterie" which included “Baker’s Latest Improved Chocolate Melangeurs, Baker’s High Class Chocolate Refining Machines, The "Climax" Patent Chocolate Coating Machine for fine goods, Patent Woodburn Fine Sugar Grinder and Disintegrator, Baker’s Chocolate Tapping Table, etc.”. In the Confectionery Pavilion, the company also showed “Baker’s Patent Automatic Sugar Wafer Machine, Patent Cream Cooling and Beating Machines, Patent Starch Moulding and printing machines, Patent Cream, Fondant and Jujube Depositing machines, the "Eureka" Patent Sugar Boiling Machine and a Patent Starch Cleaning and Tray Filling machine”. It is likely that not all of these machines were made at Willesden, indeed, Joseph Baker & Sons’ 1904 Confectionery machinery catalogue lists over 400 items of production equipment offered to the confectionery trade alone – from vacuum cookers to rolling pins.

Chocolate refiners had been made at Willesden since the eighteen-nineties and a very wide range of types and sizes were on offer. The early machines had the rolls in line horizontally, in double or treble configurations, but at the turn of the century, the five roll vertical refiner was introduced to which, in 1904, Joseph Baker & Sons fitted chilled iron rollers in place of the traditional granite rollers. There was some reluctance to accept this innovation at first but customers soon came to realize that not only could the iron rollers increase output but also obviated the overheating to which granite rollers were prone. Finding the correct metal did create some initial difficulties but Ralph Baker and, later, John Pointon and Gordon Lewis succeeded in producing a machine which was accepted as being as good as the highly efficient machines then available from Germany.

Conditions in the food industry were difficult in 1904, a big increase in the duty on sugar causing a marked depression in the confectionery and allied trades. However, orders for Joseph Baker & Sons’ equipment were fairly well maintained.

Joseph Baker & Sons’ 1910 catalogue shows a full range of machinery for processing cocoa beans – nibbing, husking, winnowing, bean sorting, cleaning and roasting. The sale of cocoa bean preparation equipment appears to have ceased at around the time of the merger with Perkins Engineers Ltd in 1920. From around this same time, the company began to drop many of the "consumables" from its catalogue to concentrate on the development of specific key unit machines and complete lines of process plant.

Shortly before WW1, an arrangement was made with John Mackintosh & Sons of Halifax to market a confectionery twist-wrapping machine developed by an employee of Mackintosh’s, Harry Bradwell, who, co-incidentally, had previously been employed by Rose Brothers. Baker Perkins manufactured quite large numbers of these machines.

INSERT - Photograph of “Champion” Patent Biscuit Coating Machine – Page 6973 – Catalogue No. 196 and The “Economic” Cooler – Page 6979.

At the time of WW1, Joseph Baker & Sons were offering the “Champion” Patent Biscuit Coating Machine – an early form of chocolate enrobing plant – in which the product was placed on a wire tray and then dipped in a bath of chocolate. After being lifted from the chocolate, the tray passed to a shaking or tapping apparatus to remove excess chocolate, the tray was then covered by a receiving tray, turned over and the coated goods taken immediately to a static cooling cupboard.

Fondant and crème, often used to produce the centres of chocolates, were produced using Joseph Baker & Sons machinery before the turn of the century. At that time basically a batch process, the company was able to provide a complete range of equipment that, over time, was developed into a semi- automated process. In 1916, Joseph Baker & Sons was offering the "Handy" Cream and Fondant Plant – "The only satisfactory Cream and Fondant Making Plant placed on the market". Syrup from the receiving pan at the left of the machine was pumped through a “Handy” cream boiling machine which heated the syrup together with glucose before being discharged onto the surface of the large water-cooled drum. This almost instantaneously cooled the film of syrup that was then passed through a continuous cream beating machine.

The company continued its policy of working with leading unit machine suppliers in the field and from December 1919, Joseph Baker, Sons & Perkins became an agent for Savy Jeanjean of Paris. This gave it access to, in particular, Savy’s chocolate enrobing technology and, through Savy, that of National Equipment of Springfield, Mass. USA, In the 1920’s and 1930’s Baker Perkins worked closely with both companies on the development of chocolate enrobing plants (see below). The Savy association also brought Bakers the agency for the Swiss-made Sapal wrapping machines. At this time the company was also selling cocoa presses from Fred S. Carver of New York.

It is interesting that after WW1, Joseph Baker Sons & Perkins catalogues showed a more lightly-built "Continental Type "5-roll vertical Refiner, described in the sales literature as - "Built in our Belgian works"– presumably to Savy Jeanjean designs.

The original “UNIVERSAL” Chocolate Enrober (a 1920 Savy Patent?) was a French invention, introduced to the trade by Savy Jeanjean. A sales brochure of the time suggested that – "It revolutionised the English language and became, indeed, a household word with chocolate manufacturers in every part of the civilized world".

The Baker Perkins years

The company exhibited at the British Empire Exhibition at Wembley in 1924 and 1925 (See also Trade Exhibitions). Again these were fully working exhibits and, in particular, the high-speed toffee and caramel plant, operated by Carsons of Bristol, caught the imagination of visitors. Based on the old Metcalfe Shaw invention, toffee in a plastic state was fed to a batch roller and then to a wrapping machine with first cut, and then wrapped individual pieces. It was clear that wrapping machines, particularly those capable of handling chocolate and sugar confectionery would be important to Baker Perkins’ future growth.

Getting into the packaging business

The Forgrove Machinery Company Ltd

The Forgrove Machinery Company Ltd had come into being in August 1901 after two young men employed in the engineering laboratories of the Yorkshire College had visited a soap works in 1897. Seeing numbers of girls wrapping bars of soap by hand, Andrew Forbes and Frederick Grover felt that a mechanical device should be provided to carry out this tedious but necessary operation and, in their spare time, they designed and made an automatic machine for this purpose. It was not a commercial success and Forbes and Grover adapted it for wrapping pieces of butterscotch and offered it to Parkinsons of Doncaster. This was not a success either but it worked admirably on oblong tablets of moulded chocolate. The "Old Rocking Horse" as it became to be known because of its action, could wrap at a hundred pieces per minute and was soon being sold to chocolate makers in England and Switzerland.

In 1903, a machine for foiling circular medallions or round croquettes was developed. Following a move to larger premises in Dewsbury Road, Leeds in 1904, the design of what became to be known as the Caramel Wrapper was begun. 1906 saw round croquette type machines being sold to foil circular medicated tablets in tin-foil from reels. The United States rights for the Caramel Wrapper were negotiated in 1907 with a Mr. Page – who was also involved in the earlier days of the Package Machinery Company – and the Consolidated Wrapping-Machine Company was formed, with Savy Jeanjean as an interested party and a number of Caramel Wrappers were made in the USA. The Forgrove Company made no lasting commercial development, however, and the Caramel Wrapper rights in the USA were sold for a lump sum.

In the next two or three years, a noisette machine was designed, developments began with the "Knott" twist wrapping machine – originally an American Patent – of which two forms, a duplex and a single–head type, were ultimately made. Known as "Oliver Twist Wrappers", they were sold until the early 1920’s, being superseded by faster machines. A labeling device for the caramel Wrapper was also evolved. Sales of twist wrappers were made in increasing numbers to Germany right up to the start of WW1.

In the years following WW1 Caramel Wrappers were in constant demand but a demand arose for twist-wrapping pre-cut pieces of toffee for which the company developed its No 22 and 22A Twist Wrappers.

Rose Brothers (Gainsborough) Ltd

William Rose patented his first wrapping machine in 1885. It was originally intended to produce packs of tobacco, which had hitherto been weighed and wrapped by hand. Three years later he sold his first machine and Rose began to apply the principle of this machine to the wrapping of other products. Within a few years, Rose Packaging machines were to be found in chocolate and confectionery, bakery, biscuit and tea factories. In 1906, William and his brother incorporated themselves as Rose Brothers (Gainsborough) Ltd.

Returning to the situation vis-à-vis caramel wrapping equipment in 1925, the Forgrove machine was the most successful, having sold seventy-six machines to Stollwerk of Cologne in that year alone. The machine had a simple electro-magnetic device to control the wrapping-paper feed instead of the mechanical system adopted by the others but Rose and Baker Perkins had moved ahead in twist wrapping. The need for Baker Perkins to have a bigger presence in wrapping and packaging led, in 1927, to an arrangement whereby Baker Perkins Inc., Saginaw would act as Forgrove agents in the USA and was followed, in 1928 with an agreement in which Baker Perkins and its associated companies, sold Forgrove machines exclusively, both at home and abroad.

Baker Perkins also took a minority holding in Forgrove. However, it was considered at the time that a trading association with William Rose was totally out of the question. In 1930, the Baker-Martin-Harris "Plastic" twist-wrapping machine was transferred, together with Mr. Harris, from Peterborough to Forgrove at Leeds.

Collaboration in Chocolate Enrober Development

In around 1929, the collaboration between Baker Perkins, Savy Jeanjean and National Equipment brought forth the successor to the “Universal” Enrober. The sales catalogue declared - “When the first patents expired, the copyist fancied that his chance had come. All he had to do was to duplicate the essential mechanism, which was now open to his hand, and the rest would be easy. Copies have, in fact, appeared in several countries. Meanwhile, however, the original inventors have not been idle. They and their associates in England and the United States of America (Baker Perkins and National Equipment) had been giving patient and continuous study to the Enrober and its problems. They knew, better than anyone, where improvements could and should be made. They determined to give to the trade an Enrober as far in advance of the original and its copies as that original had itself been ahead of earlier methods of machine-coating”. Built around cast-iron frames, this became known as the "International" Enrober.

Further collaboration with both its customers and its trading partners resulted in the Baker Perkins’ 1929 Confectionery Machinery catalogue showing a "Rowntree-Baker Conche" and a Chocolate Moulding Plant produced by Savy Jeanjean but believed to have drawn heavily on National Equipment patents. This catalogue also included batch roller and candy pulling machines.

More Ideas on Sugar Cooking

Metcalfe Shaw continued to develop his ideas on sugar cooking between the wars and developed the "Microfilm" continuous sugar-cooker in about 1930. In this, the syrup was spread in a thin film over the steam-heated inner surface of the cooking tube. Centrifugally loaded blades rotated and continuously scraped the surface at such a speed that cooking time was reduced to about eight seconds. As with the "Eureka", Shaw successfully developed the "Microfilm" to operate under vacuum conditions. This had the advantage of a second receiving hopper, also connected with the vacuum pump and condenser. When all of the syrup had fallen into the lower hopper, air was admitted and the goods discharged. The vacuum in the machine was never broken resulting in an absolutely even quality of product. The "Eureka" could be used for both hard-boiled sweets and cream and fondants.

Development of one of Forgrove’s most successful products – the Universal Twist Wrapping Machine, Type 22B – capable of handling 12 different sizes, was begun in 1931 by a team led by Alec Russell and H.E. Gregory. The following year saw an agreement being made, in collaboration with Baker Perkins, with Messrs Hansella GmbH of Viersen, Germany. Hansella’s line of sweet producing machinery linked well with "plastic" wrapping machines such as the 22B.

Ever anxious to expand their business overseas, Baker Perkins had always adopted a flexible attitude to trading in overseas markets – taking financial interests in foreign-based companies and entering into licensing agreements in addition to selling through agents. This arrangement with the German firm of Hansella, in which, under a reciprocal agreement, Hansella acted as a Baker Perkins agent in Germany and elsewhere for the sale of wrapping and packaging equipment, was a typical example. The agreement lasted until 1939 but was revived again after WW2.

Innovations in the Thirties

In the early 1930s, a Baker Perkins confectionery consultant, Clifford Clay, came up with the idea of a continuous process for crème making, by adding an attachment to a Fondant Plant to continue the fondant process into a crème for delivery direct to a Starch Moulding Plant. Unfortunately, this idea did not catch on, the most likely reason being that the industry was then not ready for continuous production as it was still producing small quantities of many different varieties of products.

By the mid 1930’s, the twist-wrapping machine taken over by Forgrove from Baker Perkins had been considerably developed and a fold-wrapping version designed. Speeds of up to 600 per minute were now possible.

Forgrove acquired, in 1935, the rights of manufacture and sale for the "Transwrap" machine from its inventor, Walter Zwoyer of Messrs. Henry Heide Inc of New York. Originally designed to produce individually wrapped sweets in heat-sealing cellulose film and "panned" sweets in heat-sealed packets, the machine was considerably developed in later years to handle a very wide range of products.

A new plastic twist-wrapping machine – the "Forsella" - was introduced by Forgrove in 1937. Designed in conjunction with Hansella, it was a combination of their sweet-forming die and a 42-type high-speed twist-wrapper. A new Chocolate Tablet Wrapping machine – the Type 55 – was designed and a prototype built in 1938/39 but did not go into production until after WW2.

The "International" Enrober launched in 1929 was later renamed the New International Enrober. It was replaced with the Baker Perkins 67CQ and 68CQ All Steel Enrober, so named by its fabricated steel plate construction which was designed just before WW2, but not produced until 1945/6.

The Introduction of Sugar Rationing

Following the declaration of WW2, rationing came into force in the UK on 8th Jan 1940. All sorts of essential and non-essential foods were rationed as well as clothing, furniture and petrol. Rationing of sweets and chocolate - 12 ounces per person per month - began on 26th July 1942.

The post-WW2 years

The Forties - Shortages, Shortages, Shortages

Gordon Steels, who joined the company just after the war and became technical manager of the division, remembers what conditions were like in the Drawing Office immediately after the War:

“In the early post WW2 war years, the Drawing Office sections and the Sales Office sections were separated with all the trading sections of the D.O. coming under the responsibility of one manager who was also a director of the company. Similarly, all the sales sections were the responsibility of one manager and director. In the first years after the war, R&D activities were at a low level or virtually non-existent as all technical resources were needed to handle the immense volume of orders as our customers’ factories, both at home and abroad, started to rebuild.

A shortage of draughtsmen was anticipated with the introduction of a training scheme for which twelve craft apprentices were recruited and sent away to Leicester in groups of four, to be trained as junior draughtsmen. Later, an interviewing committee was formed to select suitable craft apprentices at the age of between 19 and 20 years old, for a broader craft training plus periods of time in the new D.O. school before entering the main D.O.

Early post war shortages of raw materials, as industry changed production from munitions to peace-time requirements, created a few problems and involved extra D.O. work, such as the shortage of steel sheet. This was replaced wherever possible with aluminium sheet, of which large surplus stocks were available from the running down aircraft industry. Electric motors were in short supply, particularly the larger sizes. Chocolate refiners, for instance, required 40hp motors and all types of makes and of all proportions were used, some probably reconditioned from the scrap heaps. Different mounting platforms and vee belt drives had to be produced for almost every Refiner built.

Electricity became in short supply, particularly during the bitter winter of 1947. Power was rationed and all available to the company was used for the machine tools. In the offices we worked without heating or lighting, in overcoats and gloves at the drawing boards, until the light was too bad to see. Saturdays were worked to get the extra daylight working hours”.

No one knew in 1946 whether Hansella would come back into the picture. Although Baker Perkins had a license to make as well as sell Hansella equipment, they had chosen to import from Hansella prior to WW2. Increased demand following the war years meant that Forgrove, even with the help of sub-contractors, had to concentrate on putting only a selected list of products back into production and decided, in the interests of the business generally, to take up the manufacture of the Hansella range of sweet-forming machines – the Rostoplast and Seamless Tablet Machines, including dies and coolers, Horizontal and Vertical Batch Feeding machines, a Centre Filling Machine and a Sizing Unit. Forgrove launched one entirely new machine in 1946 – the Roll Wrapping and Labeling Machine. This brought together a row of circular sweets on their edges and wrapped them in waxed paper or wax-backed foil before attaching a label.

Forgrove introduced the "Plaswrap" for forming and wrapping sweets in 1949. This single-head d high-speed unit comprising a forming head, a chain conveyor, and a 42-type twist wrapper was developed form the pre-war "Forsella”" mentioned earlier. Later, when linked to the Baker Perkins Ltd continuous "Candymaker" it provided the most compact and fully automated plant for producing and wrapping solid high-boiled sweets on the market, at an output of 650 sweets per minute. The one hundredth "Plaswrap" was despatched in 1966 to the same UK sweet producer – Bensons of Bury - who purchased the first machine and it was their twentieth "Plaswrap".

The pre-war relationship with Hansella, Germany was revived in 1950 but, in 1954, Hansella wanted to begin to manufacture confectionery machinery in the United Kingdom and so the agreement came to an end.

The Fifties - Boom Time as Sugar Rationing Ends

The process of de-rationing in the UK began in 1948 but made slow progress until 1953. Then food minister Gwilym Lloyd-George made it a priority for his Department. As well as sweets, he took eggs, cream, butter, cheese, margarine and cooking fats off the ration books. He de-rationed sugar in September 1953, partly as a result of pressure from sweet manufacturers and finally ended rationing when meat was taken off the ration books in June 1954.

With the end of sugar rationing in the UK, sales of chocolate and confectionery machinery took off. This was due to not only to established sweet manufacturers renewing their equipment after the relaxation of the ban but also many new-comers to the business saw an opportunity to make quick profits by obtaining a sweet making machine and a wrapping machine. In the London area alone, several hundred of such companies sprang up only for the inevitable intense competition to force them out of business. Even the long-established companies had to limit their ranges and concentrate on improving the quality and appearance of their own specialist products. These competitive pressures, combined with difficulties of obtaining raw materials, led to many chocolate and sweet manufacturers combining or forming alliances changing the nature of demand for production equipment dramatically.

Not all equipment was replaced after the war. Much of the basic chocolate manufacturing machinery – mixers, melangeurs, refiners and conches - made by the Bakers at Willesden, had been so well designed and manufactured that chocolate makers were reluctant to part with them, even though more modern machinery might have proved more efficient and reduced production costs. Few basic developments in chocolate production had taken place in the previous fifty years although some processes had been improved, chocolate enrobing for example, unit machines had been linked together to form automated plants and bulk handling of ingredients had become commonplace.

One of the "winners" in this turbulent time was the Microfilm Cooker. Despite the undoubted advantages of Metcalfe Shaw’s 1930 development, only half-a-dozen were sold before the Second World War and it was not until the end of sugar-rationing in the UK after the war that sales took off. With the pressure on manufacturers to produce products of the highest quality in order to survive against intense competition, producers turned to the "Microfilm", described by a manager of one of the most important sweet factories in the UK as “the best natural cook ever achieved since sugar used to be cooked in open pans over the fire”, came into its own. The addition of precise metering of flavours and acids in to a continuous flow of boiled sugar resulted in a perfectly translucent sweet, the quality of which never varied. The simplicity of operation meant that one operator could supervise two "Candymaker" plants that between them turned out a ton of sweets every hour.

1950 saw the start of a new Works of 40,000 square feet on the Team Valley Trading Estate in Gateshead for Forgrove that was now employing over 700 people between its Leeds and Gateshead Works. Designed originally for the "line production" of the "Universal" twist-wrapping machine, the new factory soon became largely self-contained, with its own drawing office. In the same year, a small company was set up with offices in Cologne, Germany, to cover sales and service of wrapping, chocolate and confectionery machinery (See History of Forgrove GmbH).

Candy Depositing and Confectionery Developments Ltd

Baker Perkins first began its interest in starchless moulding in the 1950s. Gordon Steels tells us that it was a remark made by the Cadbury starchless moulding team during one of the meetings at Bourneville, that led to the beginning of an association with, and later the acquisition of, Confectionery Developments Ltd by Baker Perkins. During this meeting at Bourneville, one of the Cadburys team mentioned that they had heard that a company called C.D. Ltd. had claimed that they had a depositing plant design that could be used for starchless moulding. They had investigated this claim but found the information was inaccurate and that the flexible moulds that C D Ltd. were using, would not handle the soft confectionery range needed for boxed chocolates and were limited to hard cremes and toffee. On returning to Peterborough this news was reported immediately as a possible future threat to the starchless moulding project, and within a few days contact was made with Chris Warren, the owner of C D Ltd., and a meeting arranged at Hemel Hempstead. The resulted in an agreement between the two companies whereby BP would sell all the C D Ltd. products both in the home market and abroad, and provide technical support as and when needed.

The traditional way of producing hard candy in the 1950s was very labour-intensive requiring hot batches of cooked sugar, weighing some 60 lbs and at around 140 deg F, to be lifted manually onto a cooling table where acid, flavour and colour were added before lifting it into the batch roller feeding a rope of material to the dies of a forming head that shaped individual candy pieces. Baker Perkins patented a new fully automated process named the "Candymaker" that obviated all of the heavy lifting and handling, the whole process being controlled by one operator.

Collaboration with Sollich, Germany

In 1957 a business arrangement was made between Baker Perkins and the German company, Sollich KG of Bad Salzuflen, an association that lasted 15 years or so. The owner, Robert Sollich, had escaped from his native town of Rostock in East Germany, where he had a small confectionery business, and started a new business in West Germany. The association led to some major developments in the design of chocolate enrobers and chocolate tempering machines. This work is described in more detail in Gordon Steels’ account of post-war machinery developments below.

Also in 1957, work began at Bedewell on continuous toffee and caramel cooking. After some difficulties a satisfactory solution only to be superceded a few years later by an improved system designed by Confectionery Developments Ltd.

Forgrove began batch production of the first model of the "Flowpak" packaging machine in 1958. First introduced in 1957, it was designed at their Gateshead factory. The "Flowpak" had many applications in the confectionery industry, particularly in the production of individual chocolate coated count-lines.

The Move to Bedewell - and Back Again

In the late 1950s the confectionery part of the Chocolate & Confectionery division at Peterborough was split away from the chocolate side and moved to Bedewell, where it was joined together with members from the candy forming machine section from BP Forgrove of Leeds.

Three major projects were started at Bedewell:

The continuous Candymaker.
The continuous Butterscotch Frame Cutting Plant.
The continuous Caramel Making Plant.

After about three years or so, due to personnel leaving, the confectionery side was moved back to Peterborough and the Chocolate & Confectionery section was reformed. By this time the Candymaker Plant prototype was complete, and apart from a few modifications required to the cooling system, was ready for factory trials. The Butterscotch plant design was almost complete and some parts were already in production at Peterborough. The Caramel Plant test rig had been on trial in the customer’s factory, tests had been completed, and the project had reached the stage where the design specification had been agreed upon by the customer for the first plant to be designed, but there were still a lot of loose ends to clear up. It was not an easy time for a rapid transfer of the part-complete projects to Peterborough, with key personnel leaving and only two members of the original team agreeing to transfer to Peterborough.

In his description of the post-war development of C+C machinery at Peterborough below, Gordon Steels highlights the profit potential of spares for some of the key machines. Roasted cocoa beans are an abrasive material and significantly erode the machinery used to process them, creating a demand for replacement parts. Cocoa liquor mills provided a very lucrative spares opportunity for replacement discs. With the discs having a running life of up to 3,000 hours, the replacements generated more profit over the life of the machine than the machine itself. Chocolate refiner rolls also needed regrinding every 10,000 hours or so - about every 2 years - and new rolls were required after four or five regrinds. However, in the 1960s the manufacture of chocolate refiners ceased so that the special roll grinding machinery installed at Westwood Works could be used to produce rolls for the expanding Printing machinery business.

The Sixties - the Acquisition of Rose Brothers

Mention has been made earlier of Baker Perkins’ attempt in 1925 to have a closer association with Rose Brothers of Gainsborough and develop a greater presence in the confectionery packaging business, only to be thwarted by William Rose, the founder of the company. The death of William’s son Alfred in 1957 followed by the deaths of other senior employees, created an opportunity to re-open discussions and Rose Brothers became part of the Baker Perkins Group in 1961.

The Rose acquisition brought onto the Group more confectionery packaging machines and know-how. William Rose had devised in his early days, a machine for wrapping small chocolate tablets. John Mackintosh of Halifax brought to him, still in an experimental stage, a twist-wrapping machine for sweets. William Rose perfected it and came to an agreement with Mackintosh that allowed him to exploit the invention. He also produced a caramel wrapper – a "pocket edition" of his packet tobacco machine.

Baker Perkins Ltd took over responsibility for selling Confectionery Developments Ltd’s sweet depositing equipment all over the world. Coupled with Baker Perkins Ltd’s Microfilm cooker, the CD depositor revolutionized the production of toffee, caramel, fondant and high-boiled sweets, the key breakthrough came with the invention of a pump that would deposit measured charges of boiled sugar to an accuracy of plus or minus one per cent. As a result of this collaboration, Confectionery Developments’ export business rose from five to sixty per cent of their turnover in three years.

Starchless Moulding - the Breakthrough

The 1960s saw an attempt by Baker Perkins get rid of the messy, and potentially highly dangerous, starch moulding process that had been used since the previous century for forming individual sweets. The potential prize was huge but the problems involved were major. The key was to find a combination of material, mould design and release agent that would allow the product to be de-moulded continuously without deformation. The breakthrough came in the mid 60’s when a Baker Perkins consultant, Ted Willets, came to Peterborough excited with a new discovery he had made in his kitchen workshop. By accident he discovered that a white powder, magnesium stearate, when coated on a mould surface, would cleanly release confectionery pieces from a mould. The story of how this chance discovery led to a close relationship with Cadburys at Bourneville and the development of a whole range of starchless-moulding equipment is told by Gordon Steels in Starchless Moulding (Phase 1) below.

An agreement was made in 1966 for Forgrove to have access to the thermo-formed plastic packaging technology developed by Robert Sendler of Recklinghausen, Germany. The equipment was thought to have wide application in the confectionery industry.

Rose Forgrove is Formed

The Group’s packaging machinery activities were brought together on 1st January 1967, with the formation of Rose Forgrove Ltd. As a result, the responsibility for sales of confectionery packaging equipment was transferred form Baker Perkins Ltd to the new company. This freed Westwood to concentrate on the manufacture and sale of chocolate and confectionery process machinery.

One of Rose Forgrove’s first developments was a major re-design of the horizontal “Flowpak” which had been introduced 10 years earlier. The new "Flowpak"– the 84 Super G - was re-designed ergonomically to make it easier to operate, easier to clean, and easier to maintain.

In May 1969, the world’s first production starchless moulding plant was despatched form Westwood Works to produce Cadbury’s Roses. The new plant cut from days to 35 minutes the time to produce 86,400 sweets per hour and reduced to five from 31 the number of employed who had previously been employed on the old, starch moulding system. The plant produced three different shapes of sweets, in three different flavours – adding a strawberry jam centre to one sweet – all at the same time.

Two of the world’s largest chocolate enrobers and cooling tunnels were delivered by Baker Perkins Ltd to Mars Ltd at Slough in June 1969. Each of these enrobers with their six foot wide bands held half a ton of liquid chocolate.

The Chocolate and Confectionery Business is Closed Down

At the time that Baker Perkins Ltd was re-organised into divisions in the late sixties, it had been decided to close down the chocolate & confectionery business, or at least allow it to run down until sales ceased. All R&D work was cancelled and the personnel transferred to the newly formed Biscuit Division. The combined technical resources were then concentrated on accelerating the development of biscuit machinery, which at the time was meeting with heavy competition and loss of sales. Orders for C&C machinery did fall to a low level but never actually dried up completely.

At the end of May 1970, Baker Perkins ltd handed over to Confectionery Developments of Hemel Hempstead, the sales and manufacturing rights of, and process responsibility for, Microfilm cookers, continuous toffee plants and continuous Candymakers. Confectionery Developments were to sell these products, plus their own depositing machinery, in the UK and Ireland. Elsewhere the regional organisations of Baker Perkins International had exclusive selling rights. Baker Perkins Ltd’s biscuit division continued to design chocolate enrobers, coolers, starchless moulders and cream plants – this equipment being sold through the bakery and biscuit divisions and the regional organisations overseas.

The Seventies - Back in Business - the Acquisition of Confectionery Developments Ltd

Baker Perkins Ltd’s involvement in sugar confectionery equipment, which had virtually ceased five years earlier, was re-juvenated in 1973 with the acquisition of Confectionery Developments Ltd. of Hemel Hempstead. A new technical team was created, without diluting the biscuit technical team, Engineers were transferred from other departments as well some outside recruitment. The redesign of key machines was undertaken and biscuit creaming machines and wafer making equipment were added to the C&C responsibilities. Later still Extruders and Cereal plant became part of C&C that then was named the CCS Division, a sub division of the Biscuit C&C Division.

A new stick pack machine was introduced by Rose Forgrove in 1974. The development was to open up a completely new confectionery-packaging field for Rose Forgrove - for pocket packs of chews and toffees.

The Eighties - Westal and Chocolate Moulding

(See also History of Westal Ltd).

In 1985, Westal Ltd of Redditch was acquired by Baker Perkins. Their business was designing and building Chocolate Moulding Plants. They had started with a 300mm mould width Micro moulding plant with a facility for "One shot depositing", in which the depositing system would place a liquid centre of for example fondant cream inside a chocolate outer shell in a single depositing stroke. The plant used plastic polycarbonate moulds that were fixed at diagonal corners in a captive mould circuit. Larger 800mm wide plants were produced, mostly for chocolate bar moulding, but the diagonal mould fixings were less stable at these wider widths.

At PAKEX in early 1986, Rose Forgrove introduced the RF 365 sweet wrapping machine – for wrapping pre-made sweets at 800 pieces per minute. Also on show was a new version of the horizontal "Flowpak", which, although based on the traditional style of machine was to an entirely new concept both in design and construction.

The Merger with APV

Following the merger with APV in 1987, the Westal business was transferred to Peterborough.

POST-WAR MACHINERY DEVELOPMENT

There follows a detailed description of the developments in confectionery machinery at Baker Perkins from the post-WW2 years up until 1990, produced by Gordon Steels who joined the company just after the war and became technical manager of the division. It is suggested that such detail might prove to be useful to indicate to non-Baker Perkins people and, perhaps, future generations that the vast, high-quality range of chocolate and sugar confectionery available on supermarket shelves is there because someone developed a process – and then a machine or plant to deliver that process – for every one of the types of confection on offer. The record also indicates that the path to success was not always smooth.

To help to understand the reasons behind the developments that took place over the years, a brief description of the processes is recorded where necessary.

List of Early Post War Machines

Chocolate	Confectionery
67/68 CQ Enrobers (24" and 32" wide)	31EF Starch Moulding M/c (Master M/c)
Tunnel Coolers	20JB Microfilm Cooker
Tier Coolers	No.2 & No.3 Batch Vacuum Cookers
3 Ton Conche	21JB Vacuum Batch Cooker (Dome Type)
4 Ton Conche	6CR Fantafilla Mould Filling M/c
30 Cwt Conche	5ET Crystallising Plant
Chocolate Kettles (300,600,1000, 2000 lbs)	No. 5 Auto Cooker
12 CK Carver Cocoa Butter Press	21 EN Cream Beater
Cocoa Liquor Mills (10 CM & 11 CM)	No.1 & No.2 Fondant Crème Plants
3 Roll Refiner (Granite Rolls)	20 & 40 Gallon Dissolving Pans
5 Roll Refiner	7A & 12A High Pressure Boiling Pans
Chocolate Block Cooler	Fondant Melting Pan
Chocolate Moulding Plant (Jensen)	Colseth Truck
5CE & 7CE Melangeur	Starch Trays
Portable Biscuit Feeder

The Development of Machines and Processes from 1949 onwards

It is not easy to put the developments that have taken place over 40 years or so in a strict chronological order, so for clarity they are divided into two sections with some indication of dates whenever possible.

CHOCOLATE

3 & 4 Ton chocolate Conches

These were completely redesigned in the 1950s to replace the old pre war design, which was a motorised version of the original line shaft driven design, each of the four pots having a capacity of either 3/4 ton or one ton of chocolate. The pots had flat bottoms made of either granite or cast iron and the ends shaped in a varying upward curve to cause the chocolate mass to curl over and back on itself for aeration to remove undesirable acidic components that remain after the fermentation of the cocoa bean. Tandem rollers of either granite or cast iron mounted in a carriage were moved backwards and forwards over the pot base by a crank mechanism which was part of a central motorised gearbox, to move the mass from end to end in addition to removing the unwanted flavours, rounding off the refined sugar crystals and braking down agglomerates of cocoa powder and sugar particles.

It is interesting to record that when a BP engineer went to Japan to commission a new 4 Ton Conche, he found an identical copy alongside.

Cocoa Liquor Mills

The pre-war 10CM & 11CM Liquor Mills had two vertically mounted grinding heads, one being a roughing head comprising four 20” diameter manganese iron one piece grinding discs, with two of the discs mounted back to back on a rotor, and the remaining two discs fitted to stationary water cooled heads. The roughing discs changed the "nibs" into a rough liquid which flowed by gravity to the finishing discs below, where it was ground on a similar arrangement of larger 27” diameter discs, with a finer grinding pattern, but in this case the discs were made in eight segments. The main post war design changes were to convert the drive shafts bearings from gunmetal plummer blocks to ball and roller bearings and the floor mounted motor remounted off the machine base on a hinged platform.

In the 1960s a new higher output liquor mill was designed and given the prefix 12CM. The grinding principle remained the same as the 10/11CM machines but the arrangement was three horizontally mounted heads, one above the other, and with diversion chutes. The liquor from the upper roughing head could be directed to the lower two finishing heads in series or in parallel depending upon the fineness required. This arrangement produced up to 1200 lbs per hour of cocoa liquor, about double that of the 11CM model.

With this development came the design and production of 27” diameter one- piece finishing discs. These discs of very hard high manganese cast iron, had to be flattened after casting so that when surface ground, would clean up with grinding edges or tips about 1/32” wide over the whole plate surface.

It took quite a lot of production effort to get the one piece discs right before they became standard to the 11CM mills as well as the 12CM type. There was one later scare when the scrap rate suddenly escalated to around 80%, but it was discovered that a change of moulding sand was the cause.

The Mills provided a continuous business for replacement discs with the running life of up to 3000 hours and generated more profit over the life of the machine than the machine itself.

5 Roll Chocolate Refiners

A Refiner reduces the cocoa solids and sugar particles in the chocolate mass from around 120 microns to 20 to 30 microns, depending upon the end fineness required. Chocolate from the mixing process is fed to the refiner with a lower cocoa butter content and in a dough-like state. The remainder of the cocoa butter is added at the final conching stage. It leaves the Refiner in a drier flake form due to the increased surface area of the solids.

The Refiner has five rolls, with four rolls vertically stacked one above the other and the lower of the five rolls off set at an angle to form a trough into which the chocolate mass is fed. The rolls were geared together to run at differential speeds to tear the fibrous cocoa particles apart and crush the sugar particles. The feed roll ran at 15 rpm and the final roll at 230 rpm with the intermediate rolls at 40, 96, and 188 rpm. The 16” diameter rolls, which are cast against a chill and were extremely hard, had a 40.5” effective surface length and were ground to a fine finish and cambered to offset deflection when pulled together to provide an even pressure over the length of the roll. The cambers and roll thickness varied from the water cooled roll to roll as well as the roll, the smallest camber on the feed roll being 0.0008” and on the final roll 0.003”.

The rolls of the first ten Refiners built after the War were cast in the BP foundry. From then on all rolls were cast at the Armstrong Whitworth works in Newcastle, and later on supplied completed with their stub ends, but the finished machining was always carried out at BP.

The first two or three hundred machines designated 4CA, had the roll pressures applied through large compression springs with a system of toggle adjustments to set the roll positions. Later the roll adjustment was completely redesigned to a positive screw adjustment mechanism between each pair of rolls, and because of additional loading, the British Timkin bearings were replaced by heavy duty Skefco self-aligning roller bearings, and the prefix changed to 4CA/SA.

After many Refiners had been built, the main drive was changed from 40hp to 60hp with a top roll speed of 310 rpm. This increased the output by approximately 35% to 1200 lbs per hour, and the machine title became 6CA/SA.

The last development was to fit hydraulic adjustments to the rolls that could maintain a constant set refining pressure without the need to make adjustments. The design had to be fitted into the existing side complicated cast side frames for reasons of cost, but it was managed without too much compromise. This machine was called the 6CA/HD Refiner.

In the 1960s the manufacture of Refiners at Westwood Works was stopped so that the special roll grinding machinery could be used to produce rolls for the expanding Printing business. Up to this time nearly 700 Refiners had been built in the post war years including one order for thirty six Refiners for the Hershey Corporation of the USA, and there were a few smaller multiple orders. Rolls needed regrinding every 10000 hours or so, about every 2 years, and new rolls were required after four or five regrinds. This business was lost with the loss of the roll grinding machines.

Every roll made had a unique identification number stamped on the end and every camber was checked and approved by the drawing office from data provided by the factory inspectors, using a master graph which was scaled to 0.025” for each 0.001” of camber and scaled at 0.025” for each 2” of roll length. This produced an exaggerated curve that had to be within 0.0002” of a true curve. It is a credit to the machine shop in the factory, that it was extremely rare for the roll camber to be rejected for being outside the limits set, and the need to be returned for regrinding, and it should stand as an example of one of the highest manufacturing skills at Baker Perkins.

The meticulous records kept during the life of each roll produced, enabled us to solve the mystery of cracked rolls. A roll in the fourth position in the stack had been returned for replacement as a crack had appeared on the surface and blamed as a defective component. The roll was broken apart and the crack was revealed by discolouration to have started on the inside of the roll and had migrated through to the outside. No reason could be found at the time and records showed that one or two other rolls had failed with cracks in the past. A few months later the mating third roll was returned cracked and records showed that the crack was in exactly the same position as the fourth roll. It was deduced that both cracks had been caused by an impact at the same time, but it had taken months, evidenced by the change in discolouration, for the cracks to migrate to the surface, the third roll taking longer than the fourth roll. This was reported back to the customer who had a number of these Refiners and was as concerned as we were, and it was remembered that months before the surface crack appeared, a machine minder had accidentally jammed a scraper between the rolls with the machine running, to remove chocolate from the rod safety guard which was fitted at the entrance to the two rolls in question. So hard were the rolls, the scraper had not made any surface marking and the incident had been forgotten. The value of good records!

Chocolate Storage and Distribution Systems

There was a great demand for the supply of storage and distribution systems and heated storage tanks were designed with a range of water jacketed fittings and pipe work. Motorised vane type chocolate pumps based upon the Enrober design were produced and a considerable D.O. input was required to tailor make the equipment to suit different customer factory layouts.

Chocolate Kettles

The original chocolate Kettles, a flat-bottomed water-jacketed circular bowl fitted with stirrers, were used to melt and store chocolate. They were made in four sizes; 300, 600, 1000 and 2000 lbs capacity, and sometimes used in pairs to temper chocolate for feeding Enrobers or Moulding Plants. They were constructed in cast iron, with an inner bowl fitting inside an outer bowl to form a water jacket, and the inside bowl machined on the inside for close fitting scraper/stirrers, which were driven from below, leaving the top accessible.

The castings had a high reject rate, often flawed with areas of porosity, and had to be scrapped, sometimes in the foundry and other times after machining.

The 1000lbs capacity Kettle became the most popular size and was redesigned using rolled steel plates for an all welded construction and a simplified drive. The cost was estimated and came out more than the old cast iron design and the design was nearly abandoned to the disbelief of the designer concerned. An investigation showed that the cost of scrapping the castings at almost a 50% rate, was lost in the works overheads and not reflected in the returned works cost. The fabricated Kettle design replaced the cast iron design, with a lessons learned by all.

At this point it is interesting to note that it was a rule of the company, or at least a rule imposed by the Estimating Dept., that costs were not to be divulged to D.O. personnel, including those responsible for the new designs. This rule was rigidly enforced by the Estimating Dept, which was responsible to the Sales Director. It was a strange situation in that those responsible for the designs and therefore setting the costs, had no idea of the costs of the design.

In later years, and with increasing competition on all fronts, the costs of the final design became as important as the design itself with cost targets being set. There was no point in completing a design with all the technical costs involved, only to find that it was non-competitive price-wise necessitating the cancellation of the project or having to start the design again from scratch. Eventually the estimating of new design costs were carried out in the designs dept., during the development of a design itself. Later still, with many variations required in the different arrangements of some plants, the D.O. design estimating facility that provided the sales dept. with cost information for sales enquiries was used.

Chocolate Enrobers

67/68 CQ Design

The early post war Enrobers, the 67 CQ 24” wide and the 68 CQ 32” wide were also known as the all steel Enrobers, as they had replaced earlier designs built with cast iron frames and tanks.

It was a heavy machine built in 0.5” and 0.375” plate with lighter gauge outer covers with hinged plate doors for access to the wire belt area. It was painted chocolate brown so even new machines looked old!

The heating of the water jackets was by steam and cold water and the mass of the metal helped to stop excessive swings in temperature control, but it required the constant attention of an experienced operator to maintain some sort of control.

72 CQ Design

This was produced during the 1950s and was of a modern appearance at the time with a separate hood with sliding Perspex panels. No paint was used above the wire belt level, the light steel plate structure being dull chrome plated. Stainless steel was then not an option because of availability and cost.

Thermostatically controlled electric heaters with a water circulation system was used to heat the machine and a worm type pump was designed to produce a non pulsating flow of chocolate in an attempt to reduce aeration. Sealed ball bearings were used throughout instead grease lubricated bushes and the overall design was in keeping with the improved hygiene requirements of the day.

It was built in three belt widths, 24”, 32”, and 42. The24” and 32” wide machines were used mainly for confectionery coating and the 42” width mainly for biscuit work. It was an easier machine to control than its predecessor, but still far from perfect by later standards.

74CQ Design

In 1957 a business arrangement was made with the German company, Sollich KG of Bad Salzuflen, an association that lasted 15 years or so. The owner, Robert Sollich, had escaped from East Germany from his native town of Rostock, where he had a small confectionery business, and started a new business in West Germany.

He had two basic ideas for controlling the temper of chocolate in an Enrober. One was Circulation Control and the other Double Stream Tempering. Both systems were used in a machine that could be attached to existing Enrobers and take over the control of the chocolate in the Enrober by maintaining the degree of temper and thereby viscosity without operator interference. The machines were called Umlaufsein Temper Machinen or U Tempers. Existing Enrobers were fitted with a level overflow on the main holding tank, and tempered chocolate from the U Temper unit was pumped continuously into this tank, and any surplus returned to the U Temper unit alongside. The circulation rate was at least twice that taken away from the Enrober at any time so that at least 50% returned to the U Temper. Here untempered chocolate from the U Temper tank joined the tempered chocolate return in equal parts, the tempered chocolate acting as a catalyst to temper the whole mass. Any return chocolate in access of the 50% returning to the tempering tube automatically over spilled into the heated tank of untempered chocolate in the U Temper tank. If items to be coated stopped for any reason, then all the chocolate delivered to the Enrober would return and automatically 50% of it would overspill into the U Temper holding tank and be detempered.

In this way the condition of the chocolate circulated within the Enrober remained in a constant state, the tendency for crystals to multiply countered by the fresh incoming tempered chocolate from the U Temper unit. The system was therefore held in a state of temper equilibrium.

At the time of the first technical meeting with Sollich KG, they had produced the first Enrobers with an inbuilt tempering system based upon the U Temper system. It was agreed that BP could build the this system into a BP designed Enrober and to this end the 72 CQ design was substantially rebuilt below wire band level to incorporate the Sollich built tempering system, and the machine renamed the 74 CQ Baker Sollich Enrober. Several machines were built and worked satisfactory with most types of chocolate, except for milk chocolate with a high percentage of cocoa butter replaced by dairy fat. For this chocolate, the system needed more retention time in it to allow the modified crystals to grow. All the experience in Germany had been with cocoa butter-only chocolate and the UK milk chocolate was a new experience in tempering for everyone.

75 CQ Design

The tempering system was substantially changed to gain some sort of retention for the high milk chocolate recipes in the UK market. It was an improvement, but the Enrober market was changing with the demand for wider and higher speed machines, particularly for biscuit coating. The inbuilt tempering system that delivered chocolate direct to the flow equipment above the wire belt handled about 6000 lbs per hour. This output was sufficient for 32” Enrobers running at around 14 feet per minute but not enough for the 48” and 54” wide machines which were later direct-coupled to biscuit plants with speeds of up to 45 feet per minute.

76 CQ Design

The requirement for wider and faster Machines for the biscuit industry, plus the problem of handling some types of milk chocolate recipes, meant a further rethink of the Enrober design. Because of the limitations of output with the built in tempering system, it was decided to go back to using an external source. This meant that the circulation of chocolate within the Enrober could be stepped up to suit the wider machines as well as for increased belt speeds. The new machine was made longer to accommodate double blowers when necessary and also be fitted with twin chocolate circulation pumps. The machine had a broad specification with widths of 24”, 32”, 42”, 48”, and 54”, and speeds of up to 45 feet per minute.

Circulation temper control via an external tempering machine was incorporated but double stream tempering was abandoned in favour of a tempering system that had sufficient retention time within it to handle all known chocolate recipes.

Cross shafts had to be bigger to keep belt width deflection within limits and the flow equipment above and below the wire belt had to be designed to obtain a uniform spread of chocolate on the wider machines. It resulted in a robust machine adequate for the 48” and 54” widths, but was over engineered when reduced to the narrower widths commonly used in the confectionery industry.

76CQ Enrobers were linked direct to biscuit plants, and for half coated biscuits ran at 45 feet per minute.

78 CQ Design

Having seen the 76CQ Enrober design, Mars of UK approached BP to design a 72” wide version for their planned chocolate bar lines. We looked over a 76CQ machine that was about to be despatched and there and then decided that a design based upon the robust 76CQ design was feasible and could be stretched to a 72” belt width with some stiffening up. It would be the largest Enrober ever built and several were to be made.

Special features required by Mars included silent blowers, a chocolate circulation of about 24,000 lbs per hour, needing a pair of new design pumps and a heat exchanger to remove the heat put into the chocolate by the circulation pumps.
The driven cross shafts supporting the wire belt were increased in diameter to keep deflection to the minimum so that the wire belt remained flat for accurate chocolate coverage and the machine made longer to accommodate extra flow equipment including twin blowers.

The machine was designed to run continuously, day and night, with one 8-hour stoppage per week for cleaning and maintenance. With this project came the design requirement for wide coolers that will be dealt with later.

Model "ACS" Design

A lower cost Enrober to replace the 76CQ design at the smaller end of the range was required, and during the 1970s a new design was produced, with, for the first time, the newly formed industrial design team being involved in the styling. The main construction of the machine below the wire belt level was for the first time made un stainless steel, using a radiator type plate with one flat surface for the body of the chocolate tank which was heated by circulating hot water. Special stick on electrical heating pads were considered as an alternative form of heating to lower the costs, but were rejected as they would take to long to heat up the machine at the end of a shift to de-temper the contents. The machine was supported on two goalposts, one at each end , and the sides covered with light weight stainless steel panels, which were easily removed for access, with the upper part in Perspex.

A feature, considered to be a good selling point, was a removable lower chocolate tank, which together with the chocolate pump assembly, was mounted on casters and could be quickly removed from the side of the machine for cleaning access, a significant point in the days when hygiene was becoming a major issue. Also the electric controls were built into a free-standing control console to keep the machine in line more accessible for cleaning, but this was to prove a disadvantage in some installations with the extra space required. The machine ended up as a square box, which together with the light gauge stainless steel panels,, gave the impression from the outside of a light weight design, although the interior components were of substance.

Although a number of the machines were sold, it did not have the appearance of an Enrober, and one of the main features, the removable tank, which incidentally had been a feature in a 1930s design, did not really catch on.

"Series D" Design

With the lack of interest in the ACS design, which in spite of its appearance, performed satisfactory, it became necessary to completely redesign the Enrober for 24”, 32”, and 42” widths.

By this time, for some unknown reason, the prefix system of machine identification that had existed since the 1920s, had been abandoned and a new name to identify the design was required. As no new name was forthcoming, for reference purposes in the D.O., it was called the "Series D" Enrober. The name was not approved commercially, but in the absence of anything better, the name "Series D" stuck.

The removable tank was discontinued, but the stainless steel construction technique was used. Unlike the ACS that had no separate hood, the machine was built up to the wire belt level as a separate assembly, and topped off with a hood housing the blower unit and electrical control panel. The lower half of the hood was fitted with sloping Perspex sides to reduce reflection and make it easier to see the interior. It also took away the box like appearance of the ACS model and in general looked again like a sturdy Enrober.

It proved to be a very adaptable design and was later made wider and longer to suit special customer product specifications.

Chocolate Tempering Machine

With the decision to abandon the built in tempering unit design, a new external tempering unit was required. With the lessons learned on the need for a system with more retention for milk chocolate, the machine was designed as a single stream tempering system.

The single stream system is divided into three stages:

Stage 1: Temperature of chocolate direct from storage at around 115 deg F., reduced to 85 deg F., which is above the temperature at which the unstable Beta Prime crystal will form.
Stage 2: Pass the cooled chocolate through the stirred retention section to allow the crystal nuclei of the stable Beta form to multiply. The minimum retention time varies with the type of chocolate. Full cocoa butter chocolate crystals grow quickly whereas chocolate containing high amounts of dairy fat needs more time.
Stage 3: Heat up the chocolate to the coating temperature which is as near as the melting point of the stable Beta crystal as practical. For dark chocolate 90/91 deg F., and milk with dairy fat around 87 deg F.

The finished tempered chocolate is fed direct to the Enrober tank and mixed in with the recirculating chocolate, and any surplus returned by the overflow pump back through a heated pipeline to de-temper it on the way back to the main storage tank.

The Mark 1 design used a vertical drum type heat exchanger with a rotor of shallow scraper blades for the first stage, followed by a stirred retention section and finally a short drum type reheat section. Incoming chocolate was pumped in through the bottom and emerged out of the top of the unit as tempered chocolate. There came later a requirement for a higher capacity tempering machine, but the drum type of heat exchanger could not be increased in area without either making the machine too high or too wide, so a Mark 2 type was designed.

The tempering principle remained exactly the same as the Mark 1 type, and used the same design of retention section, but the heat exchanger sections were replaced with a flat disc design. The flat disc had raised spiral scrapers on both sides cast into the component, the spirals on the underside so arranged to move the chocolate outwardly from the centre to the edge of the disc, and the top spirals moved the chocolate back to the centre again. The discs were sandwiched between water-jacketed platens and several discs could be assembled one above another, to provide a large total surface area. The same disc and platen assembly was used for the reheat section.

This design enabled a range of outputs to be catered for by changing the number of disc and platen assemblies and changing the capacity of the retention section to suit.

Tunnel Type Coolers

In the 1960s a modern cooler was designed to replace the pre-war wood covered design. It was named the ’Turboflow Cooler’ and used downwardly pointing jets of air to cool the products. The outer covers were made in moulded plastic, reinforced with glass fibre and with the side panels easily removable for complete access to the interior. It made in 8 feet long straight sections with 4 feet long cooling modules spaced at appropriate intervals. Each cooling module contained a return air filter, centrifugal fan, and a cooling coil of varying capacities from 3 to 5 refrigeration tons.

It was made in five widths, 24”, 32”, 42”, 48”, and 54”, and could be assembled in any combination of cooling modules and 8 feet lengths.

For example, a chocolate cooler could be assembled from three cooling modules and eleven 8 feet straight sections to make a 100-foot long cooler. The first sections could be radiant cooling instead of by direct jet air so that the wet surface of the chocolate was undisturbed by the air-flow until it was partially set.

A cooler was built to cool fruit filled pies with a high concentration of cooling modules, with one module to each 8-foot straight section. Other arrangements were built to cool biscuits and cream sandwiched biscuits. Different configurations were all built from the same common components. Air from the cooling module was directed to above the belt to a plenum chamber with the downward pointing jets with part of the air-flow directed to the support ducting below the belt to cool the bases. For very long coolers of 200 feet or so, air could be directed between the belt and the support base, to float the belt on a cushion of air and reduce belt friction and therefore stretch. It was a most versatile design and many miles of coolers were built up to the end of the 1980s.

The Turboflow jet principle is a most efficient method of removing heat by an air flow, but its heat transfer efficiency cannot be used to the full when cooling chocolate. The rate at which chocolate will give up its heat energy to change from a liquid state to a solid state, depends upon the cooling efficiency, but equally important is the safe rate at which the sensible heat of the mass can be removed. The cooling of the mass increases the degree of super cooling to allow the latent heat phase to be accelerated but without reaching the temperature at which unstable crystals can form before the majority of the stable crystallisation is complete. Incorrect cooling can undo all the good temper conditions achieved in the enrobing.

In 1978/9, we were asked to design a 72” wide Turboflow cooler to go with the new 72” wide Enrobers. Special requirements in the specification included hospital quality air filters and other s refinements. The deflection problems of the extra width of the moulded top covers complete with their jet plates was solved by enclosing square wooden beams within the mouldings to act as moulds around which the glass fibre and resin was layered. It was the moulded glass fibre and resin that formed the corrugations inside the sandwiched top roof to give it the deflective strength, not the wood spacers. This is mentioned as the manufacturer missed out the wood spacers deliberately because they did not think the wood to be strong enough and had completely missed the reason for them being there! The result was top covers that bowed and a worrying time was had until it was found that the wood battens had been missed out. To correct the manufactures mistake we found a way to correct them exactly as per the original design without having to scrap them and start manufacture all over again against the clock.

Multi Tier Coolers

Multi Tier Coolers used paper or plastic plaques carried around on a continuous circuit of flat trays at 50” pitch. The trays were kept horizontal throughout the circuit that was arranged in tiers of 4, 6, 8, or 10 tracks according to the length and configuration required.

The starting point was the decorating/transfer table following the Enrober where the wet chocolate-coated goods were transferred onto the plaques as they were pulled around a knife edge. The loaded plaques were then transferred onto the flat trays and elevated to the top horizontal run of the tiers in the air cooling area and then descended row by row until the final decent to the discharge point, where the plaque was removed from the tray and passed around a knife edge with the cooled goods being stripped onto the packing conveyor.

The unloaded plaque then rejoined the tray and was conveyed back to the feed end, where it was stripped of the tray again and transferred to the decorating table. In all the plaques were transferred four times on and off the trays during one complete circuit. The multi tier design was of the late 1930s vintage and was a difficult machine to set up. The timing of the transfer points at both ends affected each other in that an adjustment at one end affected the other end, and vice versa, and incorrect timing caused damage to plaque rods and stoppages with shear pins breaking

The tier cooler was popular in the early post war years, when extensions to factory buildings or new buildings had a low priority over the repairs to bomb damaged buildings. Factories had to make do with the space available and this is where the tier cooler design came into its own. It only used 16 feet of floor space for the actual cooler part, and the space below the overhead tier cooling section was used for wrapping and packaging.

Around the 1960s, the tier cooler was redesigned with a number of detail changes and one major change. Although the plaque and tray arrangement remained unchanged, the main change was the redesign of the plaque stripping mechanism and the transportation of the empty plaques back to the feed end in such a manner, that each transfer point could be adjusted for timing independently of any other mechanism. This together with a number of other detail improvements enabled the cooler maximum speed to reach 40 feet per minute instead of 24 feet per minute for the previous model. A further change was the wooden panels being changed to flat reinforced plastic panels and so improve the hygiene aspects of the installation.

Chocolate Moulding Plant

At the time of the transfer of the Westal Ltd of Redditch business to BP in 1985, Westal was building a loose mould plant in which loose moulds were moved intermittently around a rectangular circuit from station to station, with a facility for shell moulding when required. Shell moulding is a method of producing a chocolate with a centre filling by firstly coating the inside of the mould with a shell of chocolate which when set is deposited with a liquid filling such as fondant crème, jelly, toffee etc., and finally sealed with a cap of chocolate. Unfortunately the plant was not properly tested before despatch, and the plant was beset with many problems in the mechanical handling of the moulds throughout the circuit and proved to be expensive in both time and money to correct.

Although the Micro moulding plants continued to perform well, one or two of the other plants already in service were having a few problems that needed to be resolved, one plant being short of cooling time for the type of chocolate being used, with chocolate pieces sticking in the moulds.

CONFECTIONERY

Dates of key developments

1954	Development of a Brazil nut depositing plant to replace hand dipping with the first use of the key patented pumping mechanism to handle liquid confectionery.
1957	Continuous toffee/caramel cooking.
1959	Continuous atmospheric film cooking with colour/flavour incorporation for hard candy.
1960	Automatic centre placing for coated toffee preformed centres.
1963	Continuous vacuum film cooking with colour/flavour incorporation.
1964	Starchless moulding (1st generation)
1965	Revival of continuous fondant/crème making process.
1966	High pressure die cast moulds with PTFE covering for hard candy.
1966	Flexible silicone rubber moulds for soft candy.
1968	Refinement of pumping mechanism to handle more viscous materials.
1973	Hinged moulds for centre placing.
1975	Plug moulding for confectionery cups.
1976	Simultaneous forming of soft centred deposited shell (centre filling).
1977	Confectionery cup moulding for cherry filled fondant/crème.
1978	Complete re-design of depositing plant range to include centre filling and increased plant capacities.
1979	Centre filled bar moulding.
1981	Starchless moulding (2nd generation) using high pressure die cast moulds with patented air ejection means and automatic removal of moulds from plant.
1982	Top tamping to produce double-sided deposited shapes.
1983	Higher output centre filling.
1984	Turbofilm cooker development using plate type heat exchangers.
1985	Deformable moulding for cherry, nut or preformed centre insertion.
1985	Automatic cherry feeder.
1986	Frame-type moulding plant with automatic nut feeding and chocolate depositing.
1988	Lollipop depositing - solid, centre filled and multicoloured.

Candymaker

The traditional way of producing hard candy in the 1950s was to mix 60 parts of sucrose with 40 parts of 42DE glucose, dissolve the mass to 230 deg F., leaving 20% of residual moisture, and to batch cook it under vacuum to leave 2% to 3% of residual moisture. The batch was then manhandled onto a cooling table where an acid (usually tartaric or citric), flavour and colour was added, and mixed in by kneading the hot mass, turning it over and over until it is cooled to a plastic state. The batch, weighing about 60 lbs, would then be manually lifted from the table into the batch roller feeder of the plastic forming machine, where it would be formed into a rope and fed through the dies of the forming head and shaped into individual candy pieces. Heavy manual labour was needed, working with hot sugar at around 140 deg F.

A Baker Perkins patented new process named the "Candymaker", continuously cooked the sugar mass in a Microfilm cooker at around 310/315 deg F., cooled it down to around 295 deg F. and, using a patented mixing cone, mixed in the powdered acid, liquid flavour and colour. The same mixing device spread the mass in a thin layer onto a water-cooled continuous stainless steel band. Before leaving the cooling band, the ribbon of candy, still in a plastic state, was rolled over into a rope, which was then flattened and delivered continuously to the forming plant. No manual lifting was required and the whole process controlled by one operator.

The first Candymakers had an output of about 600 lbs per hour, and four were sold to one manufacturer in the USA, and others to UK manufacturers. There was a demand for Candymakers with higher outputs, particularly in the USA, and a larger plant, Mark II, was designed using the new 27JB Microfilm cooker to produce up to 1200 lbs per hour of finished candy. Later on, still higher outputs were required, particularly to feed multiple forming lines. The first of these plants, Mark III, used twin Microfilm cookers to double the output. Later on a coil type preheater was developed to preheat the mass to about 260 deg F., before passing through a single 27JB Microfilm cooker, and an output of 2400 lbs per hour was achieved.

Discharge systems were designed to direct the rope of cooled candy off the stainless steel belt to up to four forming lines. On one installation of four Candymakers for the production of a lollypop assortment of six colours and flavours between them, each Candymaker fed two lollypop forming lines, a total of eight in all, each producing a different colour and flavour. (Of the six different lollypops, two were duplicated). This was achieved by an automatic colour and flavour change at the mixing cone that produced an immediate change in the candy on the cooling belt. When this change reached the discharge point a rotating knife cut the rope and automatic ploughs directed the candy to the appropriate forming line.

One of the difficult problems to solve was the cooling of the cooked sugar mass before the addition of acid. When acid is added to sucrose it reacts with it and produces an invert sugar, an irreversible process that becomes a mixture of dextrose and fructose. The higher the temperature and the longer the time at this temperature the greater the reaction the more invert sugars are produced.

Glucose contains a range of sugars including dextrose, one of the sugars in invert sugar. The term 42DE simply means that the various glucose sugars are evaluated as equivalent to dextrose and means that 42 parts of the glucose solids are equivalent to dextrose. To produce hard candy, the sugars must be balanced so that there are sufficient invert sugars present so as to stop the hard candy crystallising (going sugary). If too much invert sugar is present, invert sugars being hygroscopic, the candy will pick up moisture, stick to the wrapper and become misshapen.

The early solution to cooling candy prior to the addition of acid was to allow the boiling mass to flow down a sloping flat-bottomed trough, which was about 6 feet long and jacketed for low-pressure steam. The aim was to reduce the temperature by 20 deg F., before mixing in the acid. It was still a high temperature to add acid, but the mass was immediately cooled as it hit the stainless steel belt and there was insufficient time at the higher temperature for excessive process inversion to take place. However there was very little tolerance in the cooling system and another solution was sought.

Microfilm Cookers

The original 20JB Microfilm Cooker owed its design to a gentleman called Metcalf Shaw, who was a confectionery consultant to BP. The prototype was built in around 1930 and primarily consisted of a vertical 6” bore manganese bronze inner tube, with an outer steel steam jacket and capable of a working steam pressures of 150 psi. Inside the tube was a rotor running at 1100 rpm and fitted with nearly 100 hinged blades designed to swing out by centrifugal force to scrape the sugar mass on and off the inner surface. It produced an exceptionally high overall heat transfer of 400 btus/deg F. diff./sq.ft. per hour, and it took about 8 seconds for the sugar mass to cook from 230 deg F. with 20% moisture to 320 deg F., with 2.4% final moisture content as measured by the Karl Fischer method of analysis. (The USA mostly used the vacuum oven drying method of analysis which was less reliable and gave a much lower reading of around 0.5 to 1% for the same candy sample, and this caused many a misunderstanding in that some people thought that the dryer the hard candy, the more stable it is, and a reading of 0.5% to 1% sounded better than 2.4% although the measurements were for the same sugar sample).

The only major changes in the post war 20JB Microfilms were to change the inaccessible piston ring expansion joint in the top of the cooker, to an accessible stuffing box expansion joint at the bottom of the cooker, and changes to the bottom steady bearing. At around the time of the Candymaker design, a larger 27JB Microfilm cooker was designed with a 12” bore bronze inner tube, and almost double the capacity.

A solution to the cooling prior to acid addition was to cook under vacuum. The vacuum cooking of candy had been done for many years, and although the cooking part was continuous, the discharge was batch. No one had solved the problem for a continuous discharge. The first thoughts were to have a discharge of small but frequent batches to achieve an almost continuous discharge but this idea was dropped as being impractical. At the time, batch vacuum cooking was normally carried out under 26”hg of vacuum and the batch discharged candy was only suitable in viscosity for mixing by the kneading process on cooling tables.

For the Candymaker a 20 deg F. drop was all that was needed and this could be obtained by cooking under 10” hg of vacuum, and the problem of continuous discharge seemed to be possible. Eventually a gear pump was tried, but the candy mass often grained after running for less than one hour. The problem was that the saturated candy on the vacuum side of the pump suddenly became supersaturated on the atmospheric output side and during this interchange the candy mass was being stirred. Once graining had started it could not be stopped as it acted as a seed for the following candy.

After many disappointments a solution was found by using a steam jacketed gear pump, with glands capable of standing high temperatures without failing, and means by which the heat of the pump head was not transferred to the integral drive gearbox and boiled the lubrication oil. No such propriety pump was available, so a complete special pump was designed and built at BP.

The project almost ended in disaster when the prototype was placed in a customer’s factory. It was started up satisfactory and run for about one week by the drawing office Design Engineer responsible and handed over to a standby engineer to supervise and report on the performance. Within a day or so of the handover graining was reported. The Design Engineer returned and ran the Vacuum Microfilm but could find nothing wrong. He ran the cooker without trouble for a few days and retired from the site leaving it again in the hands of the standby Service Engineer. The problem was repeated twice more, each time nothing could be found at fault. The customer lost all faith in the installation which he had intended to keep and the prototype cooker was removed and the project was all but abandoned.

Many weeks later by chance, the standby engineer came to the Peterborough office, so we had yet another inquest. He was asked yet again exactly how he started up the cooker and all seemed correct. He was then asked yet again how he closed the cooker down after a production run and this seemed to be the correct procedure. He was then asked how the thoroughly the cooker was washed out and he said that he filled the syrup reservoir with "cold" water and pumped it through the system.

The penny immediately dropped. You never ever use "cold" water to clean out a cooking system, let alone a vacuum system. It leaves a sugar solution on the inside of the syrup feed pipes which can crystallise as it dries out and can be washed off by the saturated sugar solution the next day and cause graining. Hot water and steam must always be used. The project was reinstated but it was quite some time before confidence in the system was restored, and the project continued to a successful conclusion, but it was such a close thing.

The patented Vacuum Microfilm Cooker project, which had started as a solution to the problem of acid addition for Candymakers, and had been on the point of being totally abandoned because of an elementary mistake, later played an important roll in the successful link up and the eventual purchase of Confectionery Developments Ltd., and the commercial success in world wide hard candy depositing.

The Vacuum Microfilm Cooker was built into a complete unit machine, complete with syrup tank, pump, Microfilm rotor, discharge pump, vacuum pump, liquid lactic acid, colour and flavour metering system, and control panel. Many of the unit machines were built with both sizes of Microfilms, some with preheaters with outputs in excess of 2,000 lbs per hour.

Butterscotch Frame Cutting Plant

This plant was designed at Bedewell but built in the Peterborough factory. It was a mechanised version of one of the earliest methods of making candy pieces.

Liquid hard candy was poured onto cooling tables fitted with side guides and allowed to level off for about two minutes for a skin to form on the top surface. A segmented cutting frame is then lowered into the candy causing the skin to stretch and form individual shapes, and held there until the cooling was complete. The cutting frames were then removed, and the sheet of candy broken up into individual pieces.

The new continuous process followed the exact batch principle except that the continuously produced butterscotch was cooled on a water cooled stainless steel belt fitted with silicone rubber side guides and a continuous belt of stainless steel frames were lowered into the liquid butterscotch. When cooled, the continuous sheet of segmented butterscotch was broken into individual pieces by passing over a vibrating table. The scrap or rework rate from the new plant was reduced to less than 3% compared with 15% to 20% by the old table method.

Although the design was initially for butterscotch baton shapes, it was later sold to produce the traditional loaf shaped clear mints. The process was slightly different in that the frames had to be ’racked’ by lifting them in and out the candy during cooling to prevent them sticking to the candy.

The plants produced the exact traditional shapes complete with the sharp edges that were made when broken into pieces.

Years later the manufacturers became confident that the sharp edges were no longer an issue and changed over to depositing into moulds, which although the same size and shape, produced a much smoother product.

Continuous Caramel Plant

This project started at Bedewell and test rigs were made and installed in the customers’ factory. Following a series of tests, a specification was drawn up for the final plant.

The new plant started with a vertical preheater to heat up the caramel premix direct from an existing ring main, and was followed by a Microfilm cooker and carameliser. The caramelised mass then passed through a Simplex mixer where butter and flavour were added, and then spread onto a water-cooled stainless steel belt fitted with silicone rubber side guides. At the end of the plant, the continuous sheet of caramel was cut into slabs and passed on to frame cutting machines and cut into squares.

Although the specification had been approved and based upon the factory tests, the end product was not considered to be exactly the same as the batch production product. Later, following more tests, the Microfilm cooker and caramelising unit were bypassed, and the heated premix passed directly to the Simplex mixer which both cooked and caramelised the mass albeit at a lower throughput to the satisfaction of the customer. A larger Mixer/Carameliser of the Simplex type design, but with a different blade configuration was ordered from another manufacturer.

A few years later Confectionery Developments Ltd., produced a continuous caramel cooker along the same lines but with a much improved shaft less stirrer action to use with their caramel depositing plants.

Fondant / Crème Plants

The first new design Fondant Plant was produced in 1950, a modern version of the pre-war plant. A new dome type coil cooker was used, together with a stainless steel cooling drum and a twin shafted double length cream beater instead of the four-cylinder beater of the pre-war style.

Fondant is a suspension of fine sucrose crystals averaging about 4 microns in size, carried in a liquid phase containing all the water, invert sugars, glucose and dissolved sucrose. The process is one of controlled crystallisation. A typical recipe is 85% sucrose and 15% glucose that is dissolved in water and cooked to 248 deg F. to leave 12% of moisture. It is then cooled down to about 120 deg F. without movement, to a supersaturated state, and then beaten to produce the fine crystal structure. It is then stored in lined boxes for later use in crème making and other uses in cake or confectionery recipes.

To produce crème for moulding in for example starch, the fondant, called base fondant, is melted in water heated mixing pans and colour, flavourings and invertase added. Invertase is an enzyme which in time will convert some of the sucrose crystals into invert sugars, and increase the liquid phase to soften the final mass over a few days following chocolate coating or wet crystallising to make the crème airtight and prevent drying out. During the melting, the smallest crystals dissolve into the liquid phase and upon cooling, the excess dissolved sucrose will be given up to the remaining crystals and which will grow in size but not quantity. Therefore from a standard base fondant, a crème can be made with different structures from fine to coarse according to the re-melt temperature. The higher the re-melt temperature, the coarser the final structure.

In the early 1930s a BP confectionery consultant, Clifford Clay, came up with the idea of a continuous process of crème making, by adding an attachment to a Fondant Plant to continue the fondant process into a crème for delivery direct to a Starch Moulding Plant. Part of the cooked syrup, was separated from the main stream prior to the cooling drum, and delivered direct to a continuous mixer fitted at the end of the cream beater, and mixed into the fondant mass emerging from the fondant beater. Colour, flavour and invertase was added towards the end of the mixer, and sometimes frappe was added. Frappe is usually made from a solution of sucrose with egg albumen and is used to change the texture of the crème to reduce its specific density. It can also be added so that some of the water which is held back in the frappe by a colloid will give itself up to the liquid phase as the colloid breaks down.

This development, which became known as the Baker Clay process, was accidentally rediscovered in the 1960s when searching through the D.O. archives for a different subject. Out of pure curiosity, questions were asked about the process, but no one could remember the reasons why the process was forgotten or why it seemed not to have caught on. Further research in the records showed that four plants had been made, one of them in Saginaw. It was reasoned that it was possible that the absence of good temperature controls at the time may have contributed to its failure if indeed it had in anyway failed, but the most likely lack of interest was that the industry was then not ready for continuous production, and was still producing small quantities of different variety of products. It was decided to reintroduce the Baker Clay system and following preliminary design layouts, enquiries were made in the market place and they turned up a positive enquiry from Moirs of Canada, who had been using a Baker Clay plant since the 1930s and was in the need of replacement. From this start several plants were built.

Starch Moulding (Master Machines)

Starch moulding goes back a very long way. It is a very flexible process in that moulded shapes are easily changed by simply substituting another relatively inexpensive mould board, and it can mould most confections from cremes and jellies to pastilles and hard gums.

The post war design 31EF model was based on the 1920s by Savy Jean of Paris drawings and was probably originally designed by National Equipment in the USA. It used wooden trays 32” x 14.5” x 2” high including feet, which were filled with starch, usually maize starch, leveled, printed with shaped indentations, deposited with a confection such as crème, stacked about 40 trays high and stored for 24 hours. The filled trays were unstacked onto the feed end of the Master Machine, tipped onto a moving sieve and the confectionery pieces cleaned of starch and then discharged into trays for transporting to the chocolate Enrobers. The starch was sieved and returned to a sump from which the trays were filled, and the process started all over again.

The tray movements through the machine were intermittent and varied in speed from 6 to a maximum of 10 trays per minute and were hand stacked and destacked. A tray with an average of 3 lbs of confectionery centres, weighed a total of 15 lbs. The operators at each end of the machine could each handle up to 4 tons of loaded trays per hour or 32 tons in an 8 hour shift. It is a dusty process, contaminating the surrounding areas with a layer of dust. In one reported case in the USA, a dust explosion wrecked a factory. A primary explosion in the machine area, probably caused by an electrical fault, shook the rafters scattering the starch dust and immediately a secondary explosion wrecked the factory. Pictures of the twisted roof girders were published in a food magazine and brought about requests for explosion suppression devices to be incorporated within the starch machinery. No satisfactory devices were found or fitted.

In the 1960s the 31EF machine was replaced with a more modern 48EF design with the same output specification as the ageing 31EF. The initial preliminary design layouts of what was to become the 48EF machine was of a continuous tray motion system and a spiral tower type starch separation and cleaning system. Unknown at this time, the design of a high speed machine, of up to 20 trays per minute had been undertaken by the French company, Savy Jeanjean of Paris, with the intention that it be built in Peterborough as well as in France. It would not replace the 31EF/48EF market and would enable the company to sell both low and high-speed machines. Instructions were given that the 48EF design was to be a straight replacement for the 31EF design, and the initial ideas abandoned and a more conventional intermittent tray motion machine be designed, but with a number of new features including a facility for removing the 12 feet long starch sieve cleaning system from the side of the machine. Meanwhile the high-speed prototype built in France failed to come up to expectations and was subsequently abandoned. Parts for a second prototype to be built in Peterborough were stopped.

As a result of this failure, we were asked to push the speed of the 48EF design up to 14 trays per minute and later 16 trays per minute, to fill the market gap left by the failed Savy Jeanjean design. It was too much for the intermittent tray motion, which at speed caused the starch walls to collapse and produce misshapen confectionery pieces. After a few minor modifications it ran successfully at 12 trays per minute.

Starch Tray Stackers and Unstackers (Feeders)

In the 1950s, an American company called Curry built the first Tray Stacker and this made it possible to consider higher speed starch moulding plants, although its original purpose was to reduce the heavy manual labour involved in stacking trays. Following on from the 48EF design completion, a BP Stacker was designed and included an interesting design of gearbox with two single revolution clutches, one to move the trays up and onto the stack being built and the other to move the completed stack away from the stacking area, ready to be trucked to a storage space for up to 24 hours, to allow the confectionery to solidify ready for being separated from the starch.

After a successful prototype had been commissioned, a tray Unstacker or Feeder design was started. This machine unstacked the trays one by one and fed them onto the feed end of the Master Machine. It proved to be more difficult to remove trays one by one from the top of the stack than to build a stack and although many of the components were common to the Tray Stacker, the pick up carriers were very different together as well as timing differences. Many of these machines were built both for existing and new Starch Moulding Machines.

D.O.V.O. Depositing Head

This was a French invention taken up by Savy Jeanjean and was an attachment to fit onto the depositing section of existing designs of Starch Moulding Plants, to produce centre filled confectionery, such as a jam centre inside a crème shell. Centre filling using coaxial nozzles was then a known art, but the key point of this invention was that the D.O.V.O. head, fitted in the place of an existing deposit head, used the existing single pump mechanism to produce two independent pump movements by using a lost motion mechanism. This allowed one row of pistons to deposit the shell, and a second row of pistons using their movement from part of the shell piston stroke, to place a centre inside the shell in a simultaneous movement.

The main drawback to the system was that the depositor had to be stopped if an adjustment to the volume or position of the centre deposit was needed. In spite of this it was a very successful invention which sold well for several years until eventually, new centre filling depositors became available and which could be adjusted on the run for shell and centre weight and the position of the centre.

Starchless Moulding (Phase 1)

The early attempts at starchless moulding were to use rubber mats carried in existing starch trays. After storing the deposited confectionery for about 24 hours to solidify, the flexible rubber mats were removed from the trays and bent around a barrel shaped component, which expanded the rubber in two directions so that the confectionery pieces fell out. It had a very limited application with the confectionery formulation being modified to produce a more rigid product so as not to deform during demoulding.

In the mid 1960s a Baker Perkins consultant, Ted Willets, came to Peterborough excited with a new discovery he had made in his kitchen workshop. By accident he discovered that a white powder, magnesium stearate, when coated on a mould surface, would cleanly release the confectionery pieces from a mould. He had tried a number of other powders, but only magnesium stearate would cleanly release. This discovery was taken up in Peterborough and led to a series of experiments over the next few months using every available material for moulds including egg boxes and metal gauze, and other materials both rigid and flexible. Liquid release agents were tried including slab oil and liquid paraffin!

After a considerable amount of work, Norman Mountain decided that such a development could not be carried out alone, and arranged a meeting with Cadburys at Bourneville at which a full presentation was made of all the work so far carried out at BP. Cadburys showed immediate interest and it was decided there and then to form two teams, one at Cadburys who would use their experience of confectionery to further the tests and the other team at Peterborough who would design and manufacture the test rigs required. This proved to be a good arrangement and regular meetings took place between the two teams at Bourneville or Peterborough.

As experiments progressed, a liquid release agent called Myvacet, (later others were used), was tried on rigid moulds combined with air demoulding which showed good promise. Tests with the Myvacet ran in parallel with the magnesium stearate experiments, but it quickly became apparent that the application of a liquid release agent was more practical than applying a powder to a mould surface. Eventually magnesium was dropped, helped by the fact that enquiries to the British Medical Association about the use of magnesium stearate in contact with food received a non-committal answer, in spite of it being used in medical compressed tablets to assist release from the dies. The final outcome of the tests was that rigid moulds with a number of minute holes in each cavity and coated with a liquid release agent, was the practical answer.

A plant specification was drawn up and the first Starchless Moulding Plant designed and built at Peterborough. A special multi row glandless piston depositor was designed to produce a tailless deposit into the moulds which were based upon the Cadbury design of chocolate mould and were nominally 32” x 12” in a configuration of 18 x 5 impressions, making a total of 90 impressions per mould which were deposited in one go from the multi headed depositor.

To demould the confectionery units after about 12 minutes of cooling needed a heavy construction demoulding device. With an air pressure of around 30 psi, the force over the whole mould amounted to about five tons. About six or eight small air holes of 0.014” to 0.016”were drilled in each cavity, too small to become clogged with liquid confectionery and with a high resistance to stop a pressure drop across the mould if some units demoulded fractionally before the others.
The moulds were manufactured in Cadbury’s mould-making department, with the drilling of the air release holes carried out by BP. The problem of airless spraying of the mould cavities with small quantities of Myvacet was initially solved by using modified diesel injectors from a local supplier.

The prototype went into production at Bourneville producing three different confectionery centres simultaneously in three lanes of six rows. One lane produced a straight crème and the second lane a fudge. The third lane produced a jam centred strawberry crème. It made for a quite complex depositing head.

Production started almost from the first day with very little modification or adjustment required and the plant was direct coupled to an chocolate enrobing line, with the total time from the point of deposit to the finished chocolate coated pieces ready for packaging, taking about 25 minutes and no intermediate handling. Normally the production time would take at least 24 hours with the starch process, which because of the dust involved, had to be produced in an area away from the enrobing and packaging rooms. A plant for the recovery of material for misshapes was no longer required due to the almost complete reduction of rework with the starchless system over moulding in starch. The only formulation change needed was a small change to the moisture content, (around 0.5%), to allow for the fact that this amount would be normally lost to the starch. The reason that a crème can be demoulded in 12 minutes by the starchless process as compared with up to 24 hours with the starch moulding process, is that the starchless system does not distort the crème shape which remains intact once demoulded, albeit in a fairly fragile state, whereas to separate a crème from starch, the crème has to withstand vigorous sieving and brushing to remove the starch and therefore must be allowed time to become firm enough for this operation.

Following the successful performance of the prototype plant, a second plant was ordered. Later on a third plant was ordered for the production of the Cadbury Fry chocolate-coated Turkish Delight bars at Bristol.

Confectionery Developments Ltd, Hemel Hempstead

There have been other accounts as to how BP and CD got together, but the real truth is that it was a remark made by the Cadbury starchless moulding team during one of the meetings at Bourneville, that led to the beginning of an association with, and later the acquisition of, Confectionery Developments Ltd by Baker Perkins. During this meeting at Bourneville, one of the Cadburys team mentioned that they had heard that a company called C.D. Ltd. had claimed that they had a depositing plant design that could be used for starchless moulding. They had investigated this claim but found the information was inaccurate and that the flexible moulds that C D Ltd. were using, would not handle the soft confectionery range needed for boxed chocolates and were limited to hard cremes and toffee.

On return to Peterborough we reported on this news immediately, as a possible future threat to the starchless moulding project, and within a few days contact was made with Chris Warren, the owner of C D Ltd., and a meeting arranged at Hemel Hempstead. The meeting resulted in an agreement between the two companies whereby BP would sell all the C D Ltd. Products both in the home market and abroad, and provide technical support as and when needed.

One problem that then existed for C.D. Ltd., was with the acidulated hard candy depositing. The fruit flavoured hard candy could only be made in small batches and this made the product somewhat unstable with too much process inversion and therefore becoming sticky in the wrappings. This is where the Vacuum Microfilm Cooker, developed originally for the Candymaker Plants, came into its own and allowed larger depositing plants to be produced in the future.

At the time of the first meeting with C.D. Ltd., they were installed in small old bakery premises so short of space and they had to build plants in two sections. Later with increasing orders, they moved to a new factory site and later still expanded into an adjoining factory.

Confectionery Depositing/Moulding

At the time BP and CD got together, CD had produced a two-row nut dipping depositor for the manufacture of butterscotch dipped brazil nuts. Originally it was a hand dipping process, now mechanised. They were also producing a six row hard candy depositing line with forced air ambient cooling.

The original moulds were sand cast aluminium moulds with stamped copper plate inserts which were later replaced by die cast moulds with a Teflon covering. In both cases a single spring loaded central pin ejected the solidified candy that apart from an added Teflon sleeve, remained unchanged for the next 30 years or more.

Perhaps the most important invention as well as the moulds, was the glandless depositing piston assembly, completely self-contained within a sleeve and with an outlet valve and nozzle. This assembly screwed into the bottom of a heated hopper and was produced in a range of piston sizes. It proved to be a most versatile design that could handle most confectionery products.

At the very beginning of the association, and to promote sales of depositing equipment, a demonstration plant was set up in the Peterborough experimental department. At the time only the experimental Vacuum Microfilm Cooker rig was available, which was cleaned up to make it more presentable. The demonstration line was a Vacuum Microfilm Cooker feeding a CD 12 row Depositor, (Two six row moulds side by side), and a Forgrove 5 IST Twist Wrapping Machine. Each demonstration lasted about 30 minutes with over 20,000 individual sweets being produced and wrapped during each demonstration that was given each morning and afternoon for five days, and witnessed by many interested customers both from home and abroad.

A hard candy fruit flavoured sweet was produced, using buffered lactic acid in liquid form together with colour and flavour, metered into the discharge pump from which the liquid hard candy was delivered through a steam jacketed pipeline to the depositor hopper. One plant was sold immediately following the demonstration to a UK customer and was later followed by other orders. Disposal of the sweets produced was not a problem!

During the months that followed it became obvious that larger output plants were required, particularly for the USA market, and CD were encouraged to expand their designs to meet this growing interest. Fortunately the risks of expanding the design were extremely low and mainly one of strengthening cross members and drives.

Three mould wide plants followed by four mould wide plants were made and sometimes double row depositing heads were fitted. The result was that the output of the plants increased at least four fold in the following years with further increases when a staggered pattern of impressions were made within the existing mould dimensions.

Flexible silicone rubber moulds were introduced for moulding products such as toffee that could not be ejected from a rigid mould. A slight change to the deposit valve was required to prevent fat separation in confectionery with a high fat content and refrigerated cooling air was introduced. Apart from these changes the design was essentially the same as the hard candy plant.

Centre Filling Deposited Candy

Soft centred hard candy was produced on plastic forming machines for many years. For plastic forming the candy shell or outer with a moisture content of around 3% was first cooled to a plastic state before being joined by soft centre material cooked to a lower temperature and hence a higher moisture content. When the two materials came together in the rope sizer of the forming machine, there was no chance of the heat from the higher boiled candy transferring to the centre material causing it to boil.

This was the problem of centre filling deposited hard candy where the shell is deposited in a liquid form at 290 degrees F, at about 20 degrees F. below its boiling point, but hot enough to boil the soft centre material and burst it through the outer shell. It was easy to calculate the rise in temperature of the centre material as the shell gave up its heat, but it was the immediate effect on the centre surface causing it to boil before the heat was conducted through to the centre that could not be calculated.

It was believed that if the two materials could be kept thermally apart until the last possible moment and then quickly cooled by the mould, boiling would not take place and burst the shell.

To this end a crude kitchen experiment was tried at the cost of one burnt out saucepan, and the experiment indicated that it just might be possible to soft centre fill hard candy, but it was not certain. A hand operated double piston depositor with a single coaxial nozzle outlet, using materials of low thermal conductivity was quickly made, and with a degree of trial and error to hand coordinate the deposit movements, centre filled hard candy was produced, albeit in small numbers. Boiling of the centre was overcome providing the centre deposit did not leave a tail through the shell. A piston suck back device cured this potential problem.

A hands off motorised rig was then made and this successfully produced soft centred hard candy about two weeks before an Interpack Exhibition at Dusseldorf and it was decided at short notice to announce the invention with samples and a display of photographs of the deposited centre filled candy. It was a clear shell to display a blue coloured centre and it looked remarkably like a false eye!

The first centre filling hard candy depositor was bought at the exhibition by Needlers of Hull, who recognised its potential and eventually successfully marketed centre filled deposited candy under the name of "Sweet Sensations". In the next year or so several centre filling depositing plants were sold for candy with various fillings. Flexible mould plants for chocolate centred toffee were also sold, but for this product the potential for the centre boiling in the shell did not exist.

With the motorised rig available various nozzle designs were produced and various patterns of two colour cadies were demonstrated including striped candy. This helped to generate continuing interest in the candy depositing field.

Starchless Moulding (Phase 2)

With Confectionery Developments now part of BP, thoughts turned to making the CD design of moulding plants into a full starchless moulding plant using air ejection. The problem with the original plants was the costs of the moulds with the multiplicity of minute air holes in each cavity and the heavy engineering involved in the design. A solution was found and patented in which a single central hole of a similar diameter to the ejection pin hole for hard candy, was plugged with a fluted plug, the small flutes acting as air holes for demoulding.

In addition the back of the mould was fitted with a sealed cover to make an common air manifold to embrace all the mould cavities, and air applied through a small hole in the cover at the back of the mould to demould the confectionery pieces. This meant that the demoulding force on the back of the mould only amounted to a few pounds instead of the 5 tons on the original plants.

The moulds were the same injection moulding design as the CD hard candy moulds but with a flat back for fixing a cover plate which was made in one piece to the width of the plant to which four moulds were fixed. Special mouldings fitted to the ends of the common cover plate clipped the whole mould assembly onto the chain lugs of the transporting chain. This made it easy to remove moulds from the circuit and if need be change the shapes.

Two plants were supplied to Rowntrees York, for producing a selection of chocolate coated confectionery and included a facility for an automatic mould change. A further plant was supplied to Cadburys Tasmania for producing five different shapes and confections simultaneously, twenty rows across in groups of four as part of a chocolate box selection.

Special Hybrid Plants

Special single purpose plants were designed using mostly elements of existing designs together with some new innovations. One such plant produced in the 1980s resulted from a discovery made during carrying tests out on the experimental depositor, a most useful piece of equipment that was essential in the sale of many plants.

It was discovered that by depositing crème or caramel into a flexible silicone rubber mould and then the mould cavity being pushed up from below until the liquid deposit was displaced to reach the rim of the mould cavity and then released, the inside of the mould could be lined with the deposit to form a shell. The viscosity of the material was important and also the mould shape. Using this idea, a plant was produced for a USA customer for producing a chocolate-coated cherry filled crème. This product was originally a crème moulded in starch in a cup shape that after being demoulded and cleaned of starch, was fed onto a belt conveyor, part filled with a jelly, and then a cherry inserted. It was then chocolate coated in a labour intensive operation throughout.

The new plant produced the crème shell as described, deposited with a jelly and the placed a cherry into the deposit and cooled to solidify the crème and jelly with the cherry in place. As the product was captive in the mould and therefore registered, the cherry could be automatically placed. The cooled product was discharged from the mould in a regimented pattern onto the feed belt of the chocolate enrobing line.

A hard crème was used so that it would not deform with the mould as it was demoulded and the addition of invertase to the crème recipe, softened the crème inside the chocolate shell in a few days by converting some of the sucrose crystals into an invert sugar. From depositing of the crème to the finished enrobed product, took about 30 minutes.

This same design of plant could produce a variety of products such as a caramel shell with a chocolate and nut filling or any other preformed centre, all contained within a mould until solidified. Another "hybrid" plant used a flat bottomless mould in contact with a belt that was supported in turn on a slat type conveyor. Chocolate was then deposited into the mould cavities followed by nut pieces that were then tamped down into the chocolate. After cooling, the grid moulds were lifted vertically leaving the product on the belt for direct feeding to a chocolate enrobing line.

Perhaps the most important point about the depositing/moulding machines originally designed by Confectionery Developments Ltd., is the flexibility of the design principles, which later allowed so many variations to be achieved to suit the market requirements. Although some variations required the design of new elements, no prototypes were needed, as all the data required for producing a new product was found on the experimental depositing rig used initially for the hard candy centre filling experiments, and later extended to become the main product demonstration unit.

Patents

During the post war years to 1990, many patents were granted, and if available, would have served as a guide to the years in which the developments and events took place. Unfortunately to date, no records of these patents can be found and it would appear that they may have been lost or destroyed during the move from Westwood to Paston. Access to the records kept by H.M.Patents Office has been attempted, but to date without success. If eventually patent records do turn up from another source, then they can be added to these notes and help to put the history of the Chocolate and Confectionery business in the post WW2 war years to 1990 into a more accurate chronological order.

This is not the end of the chocolate and confectionery story. Baker Perkins still produce a wide range of chocolate and confectionery equipment from their factory in Paston, Peterborough – see www.bakerperkinsgroup.com